( 321 ) 



in the Nautical Almanac of 1902. The epochs were found a little 

 earlier, to wit: 



superior conjunction of 10 July, 10.'46"^-9 M.T. Grw. O-"-? earlier 

 inferior conjunction (mean 



of ingress and egress) 16 July 5"40m0 ,, „ O^^-S 

 superior conjunction 17 July 23''54"i-4 ,, „ 0^-7 



all three less than a minute. 



Now, as the conjunctions in the Kautical Almanac have been 

 calculated by the aid of Damoiseau's tables écUptiques {mokm^ixWow- 

 ance for some slight corrections indicated by Adams) the difïerences 

 must be solely due to the fact that in Damoiseau's second part the 

 main terms only of the equations and perturbations bave been taken 

 into account. 



The same tables represent as accurately the superior conjunction 

 of I on January 1, 1908, i4"4"i-2 M.T. Grw. = January 2, 2"-13™-55 

 civil time of Paris; the error amounts to 0^-07 or O^'Ol linear 

 measure only, aii arc traversed by the satellite in 0'"-5. 



(On the terms taken into account in the second part of the tables 

 of Damoiseau vide 3''^^ appendix below). 



In his letter Mr. Stanley Williams mentions another rare obser- 

 vation, made as well by himself as by the Spanish observer J. Comas 

 of Tails, (near Taragona), on 14 August 1891, to wit of the coin- 

 cidence and of the subsequent separation of the shadows of two 

 satellites on the planet. He concludes that an eclipse must have 

 taken place. These phenomena will be treated in the second part of 

 this communication. 



(9) Below follows the table which has served lor this computation. 

 The unit, the radius of Jupiter, is 18"-37. Soüillart states that he 

 found mentioned in the papers of Damoiseau that this number was 

 borrowed from Arago. According to Houzeau, Arago must have 

 made the determination by means of the double image micrometer 

 (an invention made nearly simultaneously by himself and Pearson; 

 of the latter the observatory at Utrecht possesses a specimen). 

 Particulars about these measures are not known. The number is 

 smaller than that found by other astronomers, vide for instance 

 Houzeau, p. 647— 650; See, Astron. Xachr. N". 3670 (15 Aug. 1900). 



